Converting stick welder to tack welder.

[Barry Lennox]

On Tue, 22 Feb 2005 12:48:03 GMT, "Tim Zimmerman" <Z@UCBerkeley.edu>
wrote:

I need a tack welder for joining thin plates and electronic
components. Like the tack weld you see in your NiCad battery packs.
I have no practical use for my 120v, 80-Amp stick welder so now I'll
convert it into a tack welder.

I like to get some ideas on how to make a setup that will be safe and
precise enough to do small electronic welds like the welds found on
some relays. Does this sound possible, if not can you point me to a
place to get a spot welding setup?

I believe you are following the wrong path here. You need little
voltage, but a LOT of current.

I made a very good spot welder for batteries and similar tasks from an
ex microwave oven transformer, the biggest I could find. Hack off the
HV secondary, thousands of turns of very fine wire, then I rewound it
with just 3 turns of wire, but I packed in as much 8g and 12g wire as
would fit, and paralleled all the turns.

The secondary is controlled by a SSR (Croydom CSD2410) pulsed by a
simple 555 timer circuit. It can vary from about 75-300 mSec I can
also switch in one of 3 wirewound resistors in the secondary to give
me fine control.

The electrodes must be made to suit your exact application, and some
trial and error can be expected. I first used copper and brass, but
now get much better results from a proper spot welding electrode
machined to suit my application. It was not cheap, about $11 for a
3/8" rod about 3 inches long, but it gives very good results. It had a
trade name like "Elkalloy" IIRC.

For optimum results, it is also important to control the electrode
pressure, but I find I can achieve satisfactory results by hand.

Barry Lennox

 

[mike]

Jamie wrote:

Tim Zimmerman wrote:

I need a tack welder for joining thin plates and electronic
components. Like the tack weld you see in your NiCad battery packs.
I have no practical use for my 120v, 80-Amp stick welder so now I'll
convert it into a tack welder.

I like to get some ideas on how to make a setup that will be safe and
precise enough to do small electronic welds like the welds found on
some relays. Does this sound possible, if not can you point me to a
place to get a spot welding setup?

Thanks






find your self dead microwave oven ( high power line), use the
heater tap on the transformer.

How well did this work when YOU tried it? How did YOU keep from killing
yourself on the high voltage winding?
mike
http://nm7u.tripod.com/homepage/welder.html

--
Return address is VALID but some sites block emails
with links. Delete this sig when replying.
..
Wanted, PCMCIA SCSI Card for HP m820 CDRW.
FS 500MHz Tek DSOscilloscope TDS540 Make Offer
Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
Bunch of stuff For Sale and Wanted at the link below.
MAKE THE OBVIOUS CHANGES TO THE LINK
ht<removethis>tp://www.geocities.com/SiliconValley/Monitor/4710/

 

[mike]

Steve Taylor wrote:

mike wrote:

How well did this work when YOU tried it? How did YOU keep from
killing yourself on the high voltage winding?


When I did it, I drilled the HT winding out. Brutal, but fast. To be
honest, the idea didn't work for me at all, and I built a miniature
capacitor discharge welder that DID do the job - a modest bank of old PC
power supplies yielded enough capacitors to hold "useful" amounts of
energy.

Steve

Post some details on voltage, capacitance, how'd you switch it?
electrode construction?
mike

--
Return address is VALID but some sites block emails
with links. Delete this sig when replying.
..
Wanted, PCMCIA SCSI Card for HP m820 CDRW.
FS 500MHz Tek DSOscilloscope TDS540 Make Offer
Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
Bunch of stuff For Sale and Wanted at the link below.
MAKE THE OBVIOUS CHANGES TO THE LINK
ht<removethis>tp://www.geocities.com/SiliconValley/Monitor/4710/

 

Hi. Oddly enough, you can do spot welding with an arc welder. I tried
it, and all I got was the typical mess that you would guess. Burned up
spots with no strength.

But, just because I cannot do it does not mean that it can't be done.
There are plans on Ebay:

385664469
This auction is for a set of completely illustrated plans to build a
spot weldi
ng/cutting gun that works with your arc welder for less than $50.00.

I have not tried this out, so please buy the plans and report back to
the group.

Or, you could try Eastwood's version, which is very similar, but uses
carbon electrodes. Also, please report back to the group after trying.

http://www.eastwoodco.com/jump.jsp?itemID=795&itemType=CATEGORY&iMainCat=688&iSubCat=795

Recently, I tried to do a blind spot weld with 1/8" steel. It worked
just great. So, the problem may be power, control, and excessive heat,
which the above solutions allude to. Note that this is arc welding,
not resistance welding with a low voltage rewound microwave oven
transformer.

 

[mike]

Nick Huckaby wrote:

"Martin H. Eastburn" <oldtree@pacbell.net> wrote


I experimented with a microwave oven transformer and
was able to get 400amps at 3 volts.

With only 3 volts, the resistance of the metal and any 'dirt' best be
doing zero ohms...Not much punch through voltage.

Repeatability is a BIG issue with this. A CD system tries to deliver
fixed energy. That's less dependent in path resistance.

Martin


How about 12V? Would two car batteries work?

Sure, if you had some way to turn them on/off quickly.
Be sure to use a heavy metal box to contain the battery explosion
if something goes wrong.



--
Return address is VALID but some sites block emails
with links. Delete this sig when replying.
..
Wanted, PCMCIA SCSI Card for HP m820 CDRW.
FS 500MHz Tek DSOscilloscope TDS540 Make Offer
Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
Bunch of stuff For Sale and Wanted at the link below.
MAKE THE OBVIOUS CHANGES TO THE LINK
ht<removethis>tp://www.geocities.com/SiliconValley/Monitor/4710/

 

[mike]

Martin H. Eastburn wrote:

Eric R Snow wrote:

On Tue, 22 Feb 2005 12:48:03 GMT, "Tim Zimmerman" <Z@UCBerkeley.edu
wrote:


I need a tack welder for joining thin plates and electronic
components. Like the tack weld you see in your NiCad battery packs.
I have no practical use for my 120v, 80-Amp stick welder so now I'll
convert it into a tack welder.

I like to get some ideas on how to make a setup that will be safe and
precise enough to do small electronic welds like the welds found on
some relays. Does this sound possible, if not can you point me to a
place to get a spot welding setup?

Thanks






Tim,
The reason your stick welder is not good for spot (what you call tack)
welding is because the voltage is too high and the current too low. I
experimented with a microwave oven transformer and was able to get 400
amps at 3 volts. This is done by removing the high voltage secondary
windings and replacing them with a few windings of heavy wire or even
copper bars. See other replies for links etc. for building your own. ERS

With only 3 volts, the resistance of the metal and any 'dirt' best be
doing zero ohms...
Not much punch through voltage.

Martin

FYI

Here's the voltage waveform for a Unitek 125 intoa .001 Ohm load .
http://nm7u.tripod.com/homepage/uniwvfm.jpg
mike

--
Return address is VALID but some sites block emails
with links. Delete this sig when replying.
..
Wanted, PCMCIA SCSI Card for HP m820 CDRW.
FS 500MHz Tek DSOscilloscope TDS540 Make Offer
Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
Bunch of stuff For Sale and Wanted at the link below.
MAKE THE OBVIOUS CHANGES TO THE LINK
ht<removethis>tp://www.geocities.com/SiliconValley/Monitor/4710/

 

[Steve Taylor]

mike wrote:

Post some details on voltage, capacitance, how'd you switch it?
electrode construction?

Hi Mike,

We needed to weld some exotic metals, that required CD welding. Our
welder was built in a glove box, The electrode construction was similar
to your, we modified a toggle clamp to do the job with 1/16" diameter tips.

The cap- bank was around 2200uF (10 x 220uF 400V reservoir caps) Energy
supply was a large variable O/P PSU, large because thats what we have
around. Drive was 0-400V. Welding occured at around 40V.

Discharge was effected by a very large old automobile relay , with
contacts bigger than US pennies (around 1" - like the old UK pennies)

Job was pinched in the jaws of the spotter, then the hands had to
operate two buttons simultaneously to activate the spot.

Yes, I'd have preferred to use a huge ignitron, or a hockey-puck
thyristor, but we didn't have time - this was a two day
oh-god-we-have-to-do-this-yesterday kind of thing.

We just about managed to weld molybdenum foil ~0.2mm thick, with it.

Steve

 

[Martin H. Eastburn]

mike wrote:

Martin H. Eastburn wrote:

Eric R Snow wrote:

On Tue, 22 Feb 2005 12:48:03 GMT, "Tim Zimmerman" <Z@UCBerkeley.edu
wrote:


I need a tack welder for joining thin plates and electronic
components. Like the tack weld you see in your NiCad battery packs.
I have no practical use for my 120v, 80-Amp stick welder so now I'll
convert it into a tack welder.

I like to get some ideas on how to make a setup that will be safe and
precise enough to do small electronic welds like the welds found on
some relays. Does this sound possible, if not can you point me to a
place to get a spot welding setup?

Thanks






Tim,
The reason your stick welder is not good for spot (what you call tack)
welding is because the voltage is too high and the current too low. I
experimented with a microwave oven transformer and was able to get 400
amps at 3 volts. This is done by removing the high voltage secondary
windings and replacing them with a few windings of heavy wire or even
copper bars. See other replies for links etc. for building your own. ERS


With only 3 volts, the resistance of the metal and any 'dirt' best be
doing zero ohms...
Not much punch through voltage.

Martin

FYI
Here's the voltage waveform for a Unitek 125 intoa .001 Ohm load .
http://nm7u.tripod.com/homepage/uniwvfm.jpg
mike

I'd be nervous calling it a 0.001 ohm load - but ok.

I think the connectors are exceeding that - two clamped down with bolts and the two
on spring loaded clamps.

I'd measure the Tr fro 10 to 90% point

Thanks for the waveform and idea.

Martin

--
Martin Eastburn, Barbara Eastburn
@ home at Lion's Lair with our computer oldtree@pacbell.net
NRA LOH, NRA Life
NRA Second Amendment Task Force Charter Founder

 

[billh]

<snip>

You're being too picky.
The manufacturer publishes a specified waveform for their device
under controlled conditions.
Gives you some idea of what you're up against welding battery tabs.
mike

When I was a Senior Scientist for Schlumberger ATE, to picky isn't it.
It saved
the company and customers a number of times.

Tr is rise time of a waveform - 10-90 or 20-80. Some waveforms
measured from 0 to 100
would never make the two levels - unless a static condition exists.

Math is based on the two levels.

Martin


I don't understand you at all...
The vendor published a picture of a waveform. On it, they clearly defined
a parameter Tr and quoted a number for a time associated with that
graphical representation.
I can't think of a more unambiguous way to define the parameter.

You have in your head a definition for a thing called Tr. That's where
the ambiguity resides. And it's unrelated to the picture and number
provided by the vendor. Unfortunate choice of parameter name?
Seems so in your case.
But it's still UNAMBIGUOUS!!
Picture >= 1,000 words. Math is not required.
Feel free to relabel the vendor data to suit your needs.

mike
Me do loves a good pissing contest...

--
Return address is VALID but some sites block emails
with links. Delete this sig when replying.
.
Wanted, PCMCIA SCSI Card for HP m820 CDRW.
Wanted TEK SG504, don't need the head.
FS 500MHz Tek DSOscilloscope TDS540 Make Offer
Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
Bunch of stuff For Sale and Wanted at the link below.
MAKE THE OBVIOUS CHANGES TO THE LINK
ht<removethis>tp://www.geocities.com/SiliconValley/Monitor/4710/

While the risetime typically is not the 0 -100% rise the picture is what the
manufacturer indeed says this is what the waveform should look like. Not
surprising since the waveform is presented for setup or checking purposes
and it is much, much easier to verify with respect to the indicated points
rather than finding 10% and 90% points. Unless you have a storage scope or a
modern scope with memory it is a real PITA to locate intermediate points
with the one-shot nature of the device. Also, the rise-time in terms of the
edge of the waveform may not be the real issue but rather the the energy in
the pulse which is related to the area under the curve. The 100% point as
indicated does give an partial measurement of this and the voltage drop from
the peak is specified.

Billh

 

[mike]

Nick Huckaby wrote:

"Steve Taylor" <steve@ravenfield.com> wrote in message news:421f6d52$0$8745$db0fefd9@news.zen.co.uk...


Discharge was effected by a very large old automobile relay , with
contacts bigger than US pennies (around 1" - like the old UK pennies)


You mean a starter solenoid?


Job was pinched in the jaws of the spotter, then the hands had to
operate two buttons simultaneously to activate the spot.


Are these just safety buttons to prevent electrocution?


The cap- bank was around 2200uF (10 x 220uF 400V reservoir caps) Energy
supply was a large variable O/P PSU, large because thats what we have
around. Drive was 0-400V. Welding occured at around 40V.


This CD welders (below) require 680,000 mirco F.
What's the different about yours?

http://www.powerstream.com/spot-welder.htm

Energy proportional to the square of the voltage on the caps.

Wonder how they get 1000A through their output connector?
Wonder how they get 1000A through the small wires to their
welding tweezers?
My Unitek 125WS welder has half inch diameter output terminals
and runs #2 wire to the head.
Wonder what they use to switch the caps to the load?
Would be interesting to see a graph of weld voltage and current vs time.
mike




--
Return address is VALID but some sites block emails
with links. Delete this sig when replying.
..
Wanted, PCMCIA SCSI Card for HP m820 CDRW.
FS 500MHz Tek DSOscilloscope TDS540 Make Offer
Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
Bunch of stuff For Sale and Wanted at the link below.
MAKE THE OBVIOUS CHANGES TO THE LINK
ht<removethis>tp://www.geocities.com/SiliconValley/Monitor/4710/

 

[mike]

Nick Huckaby wrote:

mike <spamme0@netscape.net> wrote in message news:421E09C1.9040604@netscape.net...

Nick Huckaby wrote:

"Martin H. Eastburn" <oldtree@pacbell.net> wrote


http://nm7u.tripod.com/homepage/uniwvfm.jpg
Sure, if you had some way to turn them on/off quickly.


Could an automotive relay switch them on/off quickly to acheive
the wave form you posted?

"Automotive Relay" is a pretty broad term.
In general, a relay can turn on arbitrarily quickly, in terms of
risetime. Turning off can be a problem if the contacts weld together.
Even if they don't weld, there are mechanical and magnetic flux time
constants that limit the minimum pulse width.
The second pulse is problematic if the contacts are all "burnt" from the
first pulse.
I expect there are relays that could do the job, but not likely they'll
be found in an automobile.

Be sure to use a heavy metal box to contain the battery explosion
if something goes wrong.


So you're saying that capacitors are less likely to explode?

Again, we'd have to be more precise about the definition of explode.
Take a new fully charged car batery. Shine up the connections.
Slam a 1" square copper bar across the contacts. Have somebody
videotape the experiment. Report back your results. It's likely that
the result can be used for my definition of explosion.
Do NOT try this at home.

Now pick a capacitor and voltage that gives you 100 Watt-Seconds.
Short it with the copper bar. Compare the results to the battery
experiment.
mike

Thanks

--
Return address is VALID but some sites block emails
with links. Delete this sig when replying.
..
Wanted, PCMCIA SCSI Card for HP m820 CDRW.
FS 500MHz Tek DSOscilloscope TDS540 Make Offer
Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
Bunch of stuff For Sale and Wanted at the link below.
MAKE THE OBVIOUS CHANGES TO THE LINK
ht<removethis>tp://www.geocities.com/SiliconValley/Monitor/4710/

 

[billh]

"Nick Huckaby" <simoung@gmail.com> wrote in message
news:1109775854.1769515ae74296d1006e4d74363bebb0@bubbanews...

"mike" <spamme0@netscape.net> wrote in message
news:4224B217.6070909@netscape.net...
The cap- bank was around 2200uF (10 x 220uF 400V reservoir caps) Energy
supply was a large variable O/P PSU, large because thats what we have
around. Drive was 0-400V. Welding occured at around 40V. - Steve

This CD welder (below) requires 680,000 mirco F.
http://www.powerstream.com/spot-welder.htm

snip
Would be interesting to see a graph of weld voltage and current vs time.
mike

I found a graph for a dual pulse CD welder here.
http://religion.p5.org.uk/cdwelder.htm

I assume that most CD welders require about 680,000 mirco F.
How's it possible that Steve can weld with only 2200uF? How does
the power supply provide the necessary current to the capacitors?




The size of the job would presumably determine how much current is required

and thus set the minimum size of the capacitors. You can also vary the
amount of charge held in a capacitor by changing the voltage. The Charge is
Q and the Total Charge in a Capacitor is Q=CxV. Charge is also related to
Current (I) by the formula Q=IxT so IxT=CxV and then I (current)= CxV/T.
This means you can increase the current capability by increasing the size of
the capacitors C and/or the Voltage V or shortening the discharge time T.
Note that these are ideal equations and the resistance of the leads and
workpiece also factors into what happens in real life. The voltage level
also has to be within safety standards. However, you can't just increase the
current by shortening the time it is applied and expect it to work. The high
current has to be present long enough to heat and fuse the workpieces.

The above is also related to charging the capacitors. By increasing the time
to charge the capacitors you can do it with relatively small amounts of
current. In this type of welding done manually that is not hard to do
because repostioning for the next weld etc gives seconds to charge the
capacitor so the power supply does not have to deliver high currents.

Billh

 

[mike]

Nick Huckaby wrote:

"mike" <spamme0@netscape.net> wrote in message news:4224B217.6070909@netscape.net...

The cap- bank was around 2200uF (10 x 220uF 400V reservoir caps) Energy
supply was a large variable O/P PSU, large because thats what we have
around. Drive was 0-400V. Welding occured at around 40V. - Steve


This CD welder (below) requires 680,000 mirco F.
http://www.powerstream.com/spot-welder.htm


snip
Would be interesting to see a graph of weld voltage and current vs time.
mike


I found a graph for a dual pulse CD welder here.
http://religion.p5.org.uk/cdwelder.htm

I assume that most CD welders require about 680,000 mirco F.
How's it possible that Steve can weld with only 2200uF? How does
the power supply provide the necessary current to the capacitors?

You snipped the answer from my previous post. Doubling the voltage
requires 1/4 the capacitance. 18V...400V, do the math.

You have at least two major problems.
You have to store the energy somewhere. The voltage and the capacitance
are the easy parts.
Second, you have to deliver all that current to the load.
You need extremely low resistance in the switch and all the path to the
weld. The simple "fix" is to just up the energy to account for the
relatively huge losses. Looks good on paper, but makes the process
extremely sensitive to the weld resistance which depends on force,
contamination, material, luck...
Tweaking it to get a good weld ain't all that hard. Getting a dozen
good welds in a row is not so easy.

My Unitek 125 switches 400 odd volts with an scr into a transformer.
400V in >> 7V out into .001 Ohm. It's not your ordinary transformer.
I was surprised how small it is. IIRC about 2" on a side.

mike



--
Return address is VALID but some sites block emails
with links. Delete this sig when replying.
..
Wanted, PCMCIA SCSI Card for HP m820 CDRW.
FS 500MHz Tek DSOscilloscope TDS540 Make Offer
Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
Bunch of stuff For Sale and Wanted at the link below.
MAKE THE OBVIOUS CHANGES TO THE LINK
ht<removethis>tp://www.geocities.com/SiliconValley/Monitor/4710/